Exploring the factor structure of a battery of neuropsychological assessments among the CENC cohort

Abstract

OBJECTIVE: The goal of the Chronic Effects of Neurotrauma Consortium (CENC) study is to explore the effects of concussions among Service Members and Veterans. A factor model was fit to selected neuropsychological measures to identify potentially useful relationships between assessments collected on CENC-enrolled participants.

METHOD: 492 post-9/11 participants with combat exposure were enrolled across four VA study sites. Participants completed assessments including concussion history, neurocognitive functioning, and self-report questionnaires. Exploratory factor analyses (EFA) using four different methods with varimax and promax rotations were used to analyse the cognitive variables. Final model selection was based on factor loadings towards simple structure.

RESULTS: The scree plot suggested the number of factors to be extracted was between 4 and 5. EFA produced a 5-factor MINRES model with promax rotation that resulted in a factor loading with variables loading on only one factor with a predefined threshold (0.40). Variables loaded on five cognition domains: list learning, working memory/executive skills, cognitive control, fluency, and memory.

CONCLUSION: These results provide reasonable evidence that data collected from the CENC neuropsychological battery can be reduced to five clinically useful factors. This will enable us to use the factors for further study of the impact of concussion on neurodegeneration.